An Adaptive Arbitration Technique with 4×4 Mesh Topology for Low Area and High Speed NOC Design
Lenin Babu Salagala1, N. S. Murthy Sharma2, G. Laxmanarayana3
1Lenin Babu Salagala*, Department of ECE, Sri Sarathi Institute of Engineering and Technology, Nuzvied, India.
2N S Murthy Sharma, Department of ECE, Sreenidhi Institute of Science and Technology, Hyderabad, India.
3G. Laxmanarayana, Department of ECE, Anurag Collage of Engineering, Ghatkesar, India.
Manuscript received on February 10, 2020. | Revised Manuscript received on February 20, 2020. | Manuscript published on March 10, 2020. | PP: 960-965 | Volume-9 Issue-5, March 2020. | Retrieval Number: E2653039520/2020©BEIESP | DOI: 10.35940/ijitee.E2653.039520
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In recent days, On-Chip Communication is a major requirement in modern systems that produce efficient communication with less complexity and high throughput. Due to heavy traffic and increasing Network-On Chip (NoC) size, the routing algorithm produces poor performances. Normally, round-robin and matrix arbiters are used in NoC for high-speed switches. In this research work, three different types of arbiter algorithms are used such as priority algorithm, Time Division Multiplexed (TDM) algorithm, and Viterbi algorithm to improve the linearity of the NOC. Initially, the priority algorithm helps to minimize traffic congestion when the arbiter is in a busy mode. Second, the fairness and Quality of Service (QoS) of the NoC structure are analyzed by the TDM algorithm. Finally, the Viterbi algorithm based error prediction process is done in an NoC structure. Due to the usage of three proposed arbiter algorithms, the area of NoC can be reduced. The packet can be reached to the destination by using the arbitration process with less loss. In this research, the proposed methodology is called as Low AreaFault-Tolerant Adaptive Arbitration based NoC architecture (LAFTAA- NoC architecture). Finally, the Field Programmable Gate Array (FPGA) performance is evaluated such as LUT, flip flop, slices, and frequency in Spartan 6 hardware. In the proposed method, 2.42% of LUT, 3.1% of flip flop, and 8.63% of slices have reduced when compared to existing work.
Keywords: Arbiter, Field Programmable Gate Array, Network on Chip, On-chip Communication, Routing Algorithm.
Scope of the Article: Routing, Switching and Addressing Techniques